Half-dose gadobenate dimeglumine versusstandard-dose gadodiamide in dynamicmagnetic resonance imaging of non-cirrhoticlivers: a retrospective intra-individual crossovercomparison

Behrang Homayoon,1 Himanshu Diwakar,2 Evgeny Strovski,1 Darshan Bakshi,3

Alison C. Harris,4 Ruedi F. Thoeni,5 Silvia D. Chang4

1Department of Diagnostic Radiology, University of British Columbia, 3350-950 W 10th Avenue, Vancouver, BC V5Z 4E3,

Canada2Queensland Diagnostic Imaging, 457 Wickham Terrace, Spring Hill, Brisbane, QLD 4000, Australia3Peter Lougheed Centre, University of Calgary, 3500 26 Ave NE, Calgary, AB T1Y 6J4, Canada4Department of Radiology, Vancouver General Hospital, 899 West 12th Avenue, Vancouver, BC VZ5 1M9, Canada5Department of Radiology & Biomedical Imaging, University of California, San Francisco, 505 Parnassus Avenue, M-391,

San Francisco, CA 94143-0628, USA

Abstract

Purpose: The purpose of this study was to compareenhancement characteristics of half-dose gadobenatedimeglumine (0.05 mmol kg-1) with standard-dose gad-odiamide (0.10 mmol kg-1), in the assessment of hepaticvessels and lesions, using retrospective intra-individualcrossover comparison methodology.Methods: Ethics committee approval was obtained.From 2004 to 2012, 21 patients underwent MRI exam-ination with both standard-dose gadodiamide and half-dose gadobenate dimeglumine, using the same liver MRIprotocol at 1.5 T. Eighteen patients whose scans showedno artifacts were selected. Quality of liver lesion [12hemangiomas, 7 focal nodular hyperplasias (FNHs)] andliver vessel enhancement, and the global diagnosticquality of studies were ranked on a scale of 1–4 by twoindependent radiologists. Contrast-to-noise ratio (CNR)and % enhancement of liver vessels and lesions werecalculated based on region of interest, signal intensity,and noise standard deviation measurements performedat 0, 20 s, 1, 3, and 5 min post-contrast injection.Qualitative and quantitative results were compared using

the paired Wilcoxon signed rank and Student’s t-tests,respectively.Results: No qualitative differences were noted inenhancement of liver vessels, hemangiomas, and FNHs.There was no statistically significant difference betweenthe global diagnostic qualities of scans performed withthe two contrast agents. Quantitatively, liver vessels andhemangiomas did not demonstrate statistically signifi-cant differences in contrast enhancement. At 20 s, FNHsachieved higher CNR (P = 0.02) with gadodiamide.Conclusion:Half-dose gadobenate dimeglumine results insimilar contrast enhancement compared to standard-dose gadodiamide in assessment of liver vessels, heman-giomas, and FNHs, and is a reasonable alternative tostandard doses of extracellular agents in dynamic liverMRI.

Key words: Gadobenate dimeglumine—MultiHance—Dynamic liver imaging—Nephrogenicsystemic fibrosis

With advances in MRI hardware and software, and withthe introduction of hepatobiliary contrast agents, MRIis increasingly used in detection and characterizationof focal hepatic lesions and in assessment of hepatic

Correspondence to: Behrang Homayoon; email: behrangh@alumni.ubc.ca

ª Springer Science+Business Media New York 2014

AbdominalImaging

Abdom Imaging (2014)

DOI: 10.1007/s00261-014-0123-z

vasculature [1–5]. Advances in surface coils and parallelimaging, and availability of larger gradients, have re-sulted in improved image quality and higher speed ofimage acquisition which have in turn led to significantimprovements in dynamic liver imaging [2, 5].

Extracellular contrast agents are the best documentedand most widely used contrast agents in liver magneticresonance imaging (MRI) [6]. They allow assessment ofabdominal vessels and improve detection and charac-terization of liver lesions compared to non-contrast MRI[7]. The ability of extracellular agents to provide contrastenhancement is determined by the differential blood flowbetween liver parenchyma and vessels or lesions [15].Gadodiamide (OmniScan, GE Healthcare) was theextracellular agent of choice at our institution prior to2006, but its use was discontinued after several reportsassociated this agent with the development of nephro-genic systemic fibrosis (NSF) in patients with renal fail-ure [8–10].

From 2006 to 2012, gadobenate dimeglumine (Gd-BOPTA, MultiHance, Bracco Diagnostics) was used forliver MRI at our institution, as part of a strategy tomitigate the risk of NSF. Gadobenate dimeglumine is asecond generation gadolinium-based contrast agent thatdemonstrates high thermodynamic stability and hasnever been associated with an unconfounded case ofNSF [8, 11–14]. Unlike extracellular agents, which areexcreted by the kidneys, 3–5% of the administered doseof gadobenate dimeglumine is taken up by functioninghepatocytes and is excreted in bile by the canalicularmultispecific organic anion transporter, and theremaining 95–97% is eliminated by the kidneys throughpassive glomerular filtration [19–22]. This contrast agentdemonstrates twofold higher T1 relaxivity in blood(9.7 ± 0.4 L mmol-1 s-1 at 1.5 T) compared to extra-cellular contrast agents (4.3–5.0 L mmol-1 s-1 at 1.5 T),due to ‘‘transient and weak’’ interactions of its contrast-effective moiety with serum albumin [21–24].

The twofold higher T1 relaxivity of gadobenate di-meglumine led to the idea that a half dose of this agentcould result in similar contrast enhancement comparedto full doses of extracellular agents that possess lower T1relaxivities. Several studies have shown that a half doseof gadobenate dimeglumine (0.05 mmol kg-1) results incontrast enhancement that is comparable to a standarddose of gadopentetate dimeglumine or gadobenate di-meglumine (0.10 mmol kg-1) in the assessment of focalhepatic lesions [15–18].

Use of a half dose of gadobenate dimeglumine insteadof standard doses of extracellular agents is clinicallysignificant, in that it reduces the cumulative doses ofadministered contrast and potentially reduces the risk ofNSF in patients with renal failure who undergo liverMRI. To the best of our knowledge, no study to date hascompared the enhancement characteristics of half-dosegadobenate dimeglumine to an extracellular agent for

assessment of hepatic vasculature, nor have there beenany studies comparing contrast enhancement character-istics of a half dose of gadobenate dimeglumine to astandard dose of gadodiamide in liver MRI in the samepatients. Using an intra-individual crossover comparisonmethodology, we retrospectively compared liver vesseland lesion enhancement characteristics of half-dosegadobenate dimeglumine and standard-dose gadodiam-ide in our patient population.

Materials and methods

Patient selection

Following the ethics committee approval, an institutionaldatabase search from 2004 to 2012 demonstrated thattwenty-one patients underwent MRI examination at ourinstitution with both standard-dose gadodiamide andhalf-dose gadobenate dimeglumine, without havinginterval surgical or interventional therapy for liver dis-ease. Patient consent was waived by the ethics commit-tee. Three patients, in whom at least one examinationwas degraded by motion artifact resulting in compro-mised assessment of liver vessels and lesions, were ex-cluded from the study.

Eighteen patients (7 women and 11 men; age at firstscan: 53.2 ± 12.8) who met the inclusion and exclusioncriteria were selected. A total of 19 focal hepatic lesions[12 hemangiomas, 7 focal nodular hyperplasias (FNHs)]were detected in seven patients. Due to the benignity ofthese lesions, subsequent histopathologic studies werenot performed, and the diagnoses were establishedthrough clinical and radiological follow-up consisting ofultrasound, CT, or MRI examinations, with all lesionsdemonstrating stability for at least one year. Patientswho underwent therapy for liver disease after the initialMRI examination, all of whom had malignant focalhepatic lesions, were not included, as interval therapymay have led to hemodynamic alterations in the liver,resulting in altered signal intensity measurements.

The mean time interval between the MRI scans per-formed with standard-dose gadodiamide and half-dosegadobenate dimeglumine, and therefore the minimummean clinical and radiological follow-up, was18 ± 2.8 months. The study population included pa-tients with non-cirrhotic livers; therefore, none of thepatients were subject to progressive liver fibrosis and itssequelae. One of the patients demonstrated fatty infil-tration of the liver with focal fatty sparing which wasrelatively unaltered between scans. Visual inspection ofpaired MRI scans did not reveal a significant intervalalteration in study subjects‘ body habitus.

All hemangiomas showed a typical enhancementpattern characterized by an early incomplete ring ofperipheral puddling and centripetal progression of con-trast enhancement on subsequent portal venous andequilibrium phases. All FNHs demonstrated homogeneous

B. Homayoon et al.: A retrospective intra-individual crossover comparison

enhancement in the arterial phase and were isointense toslightly hyperintense on the equilibrium phases. Theywere differentiated from hepatic adenomas, either by thepresence of a central scar (n = 3) and/or enhancement(n = 5) in the hepatocyte phase of imaging on follow-upMRI examinations.

MRI protocol

MRI imaging was performed at 1.5 T (Signa Horizon,GE Healthcare) using a body-phased array coil. Eachexamination consisted of a T2-weighted Fast Spin Echo(FSE) sequence, T1-weighted in- and opposed-phaseFast Multiplanar Spoiled Gradient Echo (FMSPGR)sequences, as well as pre- and post-contrast T1-weighted2D Gradient Echo (GRE) sequences during the arterial(20 s post-injection), portal venous (1 min post-injec-tion), and equilibrium (3 and 5 min post-injection) pha-ses of contrast enhancement.

All sequences were performed with breath hold, andthe average combined acquisition time for the pre- andpost-contrast T1-weighted 2D GRE sequence was 7 minand 7 s. The T1-weighted 2D GRE axial acquisitionswere obtained with a TE of 4.2 ms, a TR of 150 ms, a flipangle of 70�, and slice thicknesses of 6–8 mm. The matrixsize was 256 9 160, and the field of view was 32–40 cm.

A power injector was used to administer either astandard dose of gadodiamide (0.10 mmol kg-1) or ahalf dose of gadobenate dimeglumine (0.05 mmol kg-1)at an injection rate of 2 ml s-1. A saline push of 20 ccwas used after contrast administration. Scanner hard-ware, imaging protocol, and image viewing softwarewere not altered during the study period.

Qualitative assessment

Qualitative evaluation of each scan was performed inmatched pairs, by two fellowship-trained abdominalradiologists with 5 and 10 years of experience inabdominal MRI. Both radiologists evaluated the qualityof all sequences, had access to the study indication andwere able to review prior imaging, but were blinded tothe contrast agent used in each examination. The qualityof liver vessel and lesion enhancement as well as theglobal diagnostic quality of each examination was thenranked on a quality grading scale of 1–4 (1 = poor,2 = satisfactory, 3 = good, and 4 = excellent), where apoor quality represents an inability to assess liver vesselsor detect and characterize focal hepatic lesions, and theremaining quality scores represented an increasing abilityby the readers to conduct detailed assessment of livervessels and focal hepatic lesions.

Assessment of global diagnostic quality was based onthe overall image quality of the study. Apart from mo-tion artifact, no other cause for image degradation waspresent. As highlighted earlier, three patients in whom at

least 1 examination was degraded by motion artifact(received a quality score of 1 by at least one reader)resulting in compromised assessment of liver vessels orlesions were excluded from the study, and all scans weredeemed at least satisfactory (quality score of 2 or higher)by both readers.

Quantitative assessment

Quantitative assessment of liver vessel and lesionenhancement was based on region of interest (ROI),signal intensity, and noise standard deviation measure-ments. Circular or elliptical ROIs were placed on eachstructure to measure its signal intensity, and care wastaken to avoid adjacent vessels and enhancing lesions.ROI noise standard deviation measurements were per-formed at the top corners of images outside of theabdomen in all scans, and care was taken to make surethat the ROIs were approximately the same size in allpatients. All signal intensity and noise standard deviationmeasurements were performed twice, and average valueswere used in subsequent calculations.

Signal intensity measurements of the liver parenchymawere performed on pre-contrast and all post-contrastT1-weighted GRE sequences at the same peripheral loca-tion in segments 5 or 6. Signal intensity of the commonhepatic artery was measured based on ROIs that coveredapproximately 75% of arterial diameter at the same loca-tion on pre-contrast and arterial-phase (20 s) post-con-trast T1-weighted GRE sequences. Similarly, portal veinsignal intensity measurements were performed on the pre-contrast and portal-venous-phase (1 and 3 min) post-contrast T1-weighted GRE sequences, and hepatic veinand intra-hepatic inferior vena cava (IVC) signal intensi-ties were performed on the pre-contrast and equilibrium-phase (3 and 5 min) post-contrast T1w GRE sequences.

Signal intensity measurements of focal hepatic lesionswere performed on areas of maximal enhancement whichwere initially identified through visual inspection of allpost-contrast sequences and confirmed by serial ROImeasurements. The highest attainable signal intensitymeasurement was selected as the area of maximalenhancement. The same area within the lesion was iden-tified on the pre-contrast, and the remaining post-contrastT1-weighted GRE sequences and signal intensities wererecorded using ROIs of approximately the same size.

The signal intensity and noise standard deviationmeasurements were then used to calculate liver lesionand vessel contrast-to-noise ratios (CNR) as a measureof conspicuity, as well as % enhancement as a measure ofrelative enhancement, based on the following formulas:

Contrast-to-noise ratio CNRð Þ

¼ Structure SI tð Þ � Liver SI tð Þð ÞNoise SD tð Þ

B. Homayoon et al.: A retrospective intra-individual crossover comparison

% Enhancement

¼ 100 � Post-contrast SI tð Þ�Pre-contrast SI tð Þð ÞNoise SD tð Þ :

Signal intensity is denoted by SI, and timing of theimage acquisition is denoted by t.

Statistical analysis

CNR and % enhancement calculations obtained witheach contrast agent were compared using the pairedStudent’s t-test. Qualitative results recorded by bothradiologists were evaluated using the paired Wilcoxonsigned rank test. A P value of less than 0.05 (P < 0.05)indicated a statistically significant difference.

Results

Qualitative results

The results of the qualitative assessment are shown inTable 1. Reader A ranked all liver lesions and vessels aseither good (quality score = 3) or excellent (qualityscore = 4), and Reader B ranked them as either satis-factory (quality score = 2) or good (quality score = 3).No statistically significant qualitative differences werenoted in the readers’ ability to assess liver vessels or de-tect and characterize focal hepatic lesions.

Reader A did not demonstrate a preference for eitherof the contrast agents when assessing overall diagnosticquality. Reader B demonstrated a preference for half-dose gadobenate dimeglumine over standard-dose gad-odiamide, but this difference did not reach statisticalsignificance (P = 0.20).

Quantitative results

As demonstrated in Table 2, the common hepatic artery,portal veins, hepatic veins, and intra-hepatic IVC did notdemonstrate any statistically significant differences inCNR or % enhancement, in their respective post-contrastphases of maximal enhancement.

Table 3 demonstrates that in the arterial phase ofenhancement, FNHs have higher CNRs (P = 0.02) withstandard-dose gadodiamide, but they do not show anysignificant quantitative differences in the remainingphases of contrast enhancement. Differences in the CNRand % enhancement of hemangiomas did not reach sta-tistical significance.

Discussion

Patients with decreased kidney function, and a concur-rent pro-inflammatory condition, are susceptible to NSFonce they receive gadolinium-based contrast agents [26].To reduce the incidence of NSF, most institutionshave introduced policies aimed at limiting the use of

Table 1. Qualitative assessment of liver vessels, hemangiomas, and FNHs, and the global diagnostic quality of MRI scans based on median qualityscores

Parameters Reader (A, B) Median quality scores (1–4)

Gadobenate dimeglumine Gadodiamide P value

Vessels CHA A 3 3 0.63B 2 2 0.70

MPV A 3 4 0.35B 3 2 0.23

RPV A 4 4 0.12B 3 3 1

LPV A 4 4 0.79B 3 2 0.77

RHV A 4 4 1B 3 3 0.79

MHV A 4 4 1B 2 2 0.79

LHV A 4 4 0.67B 2 2 1

IVC A 4 4 1B 3 3 0.18

Lesions Hemangiomas A 4 3 0.30B 3 3 0.48

FNHs A 3 3 1B 2 3 1

Global diagnostic quality A 3 3 0.78B 3 2 0.20

Quality Scores (1 = poor, 2 = satisfactory, 3 = good, 4 = excellent); CHA common hepatic artery,MPV main portal vein, RPV right portal vein,LPV left portal vein, RHV right hepatic vein, MHV middle hepatic vein, LHV left hepatic vein, IVC intra-hepatic inferior vena cava, FNH focalnodular hyperplasia

B. Homayoon et al.: A retrospective intra-individual crossover comparison

gadolinium-based contrast agents [8, 11–14, 25]. Thesepolicies include minimizing the use of these agents inpatients with renal dysfunction, decreasing the use ofless-stable agents in favor of the more-stable ones, andreducing the total administered doses of contrast inpatients at risk of developing NSF [11]. Use of half-dosegadobenate dimeglumine can be a critical component ofthis strategy with the added benefit of reducing the costof contrast administration. In a retrospective single-institution study, Martin et al. determined that inpatients with end-stage renal disease who underwentcontrast enhanced MRI, the incidence of NSF was sig-nificantly reduced from 2.6% in those who had receivedroutine doses of gadodiamide to 0% in those who hadreceived half-dose gadobenate dimeglumine [8].

To the best of our knowledge, our study is the first toemploy an intra-individual crossover comparison meth-odology to compare the enhancement characteristics ofhalf-dose gadobenate dimeglumine to a standard dose ofan extracellular agent for assessment of hepatic vascu-lature. As depicted in Tables 1 and 2, no statistically

significant qualitative or quantitative differences werenoted between examinations performed with a half doseof gadobenate dimeglumine compared to those per-formed with a standard dose of gadodiamide.

Furthermore, apart from a single statistically signifi-cant quantitative difference in the CNR of FNHs in thearterial phase of enhancement (P = 0.02), no other sta-tistically significant differenceswere identified between thetwo contrast agents (Table 3). The higher CNR of FNHsin the arterial phase of enhancement with standard-dosegadodiamide did not translate into a qualitative difference(Table 1) nor did it hinder the readers’ ability to detect andcharacterize FNHs. Moreover, there were no statisticallysignificant differences in the % enhancement of FNHs inthe arterial phase. Dynamic liver MRI images afterintravenous administration of a half dose of gadobenatedimeglumine and a standard dose of gadodiamide in threedifferent patients with and without focal hepatic lesionsare shown in Figures 1, 2, and 3.

Several studies have also shown that use of half-dosegadobenate dimeglumine in dynamic liver magnetic

Table 2. Quantitative assessment of vascular structures based on CNR and % enhancement calculations

CNR % Enhancement

Vessel Time Gadobenate dimeglumine Gadodiamide P value Gadobenate dimeglumine Gadodiamide P value

CHA 20 s 23.50 ± 14.84 23.69 ± 15.34 0.97 183.53 ± 62.16 182.36 ± 91.95 0.97MPV 1 min 26.48 ± 9.10 30.15 ± 15.91 0.37 199.64 ± 59.96 248.20 ± 96.19 0.08

3 min 27.85 ± 11.90 29.63 ± 13.58 0.67 187.93 ± 65.83 228.89 ± 83.49 0.07RPV 1 min 21.44 ± 10.64 23.82 ± 14.25 0.50 191.00 ± 63.18 190.26 ± 62.66 0.96

3 min 21.09 ± 12.04 22.98 ± 14.43 0.67 170.54 ± 54.32 175.58 ± 59.86 0.75LPV 1 min 25.15 ± 12.13 26.22 ± 15.68 0.80 182.74 ± 60.16 182.74 ± 64.89 0.20

3 min 22.88 ± 11.72 28.44 ± 17.16 0.24 158.31 ± 45.77 168.31 ± 57.91 0.20RHV 3 min 13.26 ± 13.20 17.78 ± 13.29 0.26 188.97 ± 85.22 201.32 ± 69.05 0.51

5 min 12.43 ± 11.76 12.84 ± 16.35 0.91 177.53 ± 67.71 173.97 ± 83.10 0.85MHV 3 min 13.59 ± 13.17 18.98 ± 13.48 0.11 164.81 ± 42.91 205.18 ± 150.38 0.26

5 min 10.52 ± 11.61 15.96 ± 14.78 0.27 149.08 ± 43.46 181.00 ± 134.07 0.27LHV 3 min 13.74 ± 9.92 17.75 ± 10.86 0.15 160.64 ± 43.17 167.78 ± 43.29 0.64

5 min 8.73 ± 11.13 14.66 ± 13.33 0.73 137.73 ± 45.94 142.57 ± 40.74 0.75IVC 3 min 15.34 ± 9.80 22.46 ± 12.76 0.07 152.66 ± 71.02 181.35 ± 76.29 0.07

5 min 16.16 ± 12.07 18.36 ± 13.08 0.49 150.17 ± 52.35 157.21 ± 43.33 0.67

CHA Common hepatic artery, MPV main portal vein, RPV right portal vein, LPV left portal vein, RHV right hepatic vein, MHV middle hepaticvein, LHV left hepatic vein, IVC intra-hepatic inferior vena cava, FNH focal nodular hyperplasia

Table 3. Quantitative assessment of hemangiomas and FNHs based on CNR and % enhancement calculations

CNR % Enhancement

Lesion Time Gadobenate dimeglumine Gadodiamide P value Gadobenate Dimeglumine Gadodiamide P value

Hemangioma 20 s 8.05 ± 16.08 12.85 ± 19.02 0.28 87.37 ± 62.61 110.38 ± 73.34 0.321 min 13.71 ± 14.79 17.03 ± 17.65 0.13 124.58 ± 68.89 120.88 ± 94.86 0.843 min 17.78 ± 12.61 9.05 ± 22.62 0.09 122.52 ± 59.77 119.62 ± 92.98 0.915 min 14.83 ± 15.52 8.21 ± 18.38 0.20 112.95 ± 68.91 112.61 ± 84.67 0.81

FNH 20 s 25.22 ± 8.05 34.58 ± 9.97 0.02* 95.96 ± 25.13 76.15 ± 25.66 0.241 min 24.31 ± 12.30 22.45 ± 10.45 0.78 90.11 ± 26.48 83.17 ± 25.93 0.703 min 31.13 ± 11.98 24.26 ± 10.27 0.32 87.07 ± 26.59 71.53 ± 20.86 0.385 min 24.00 ± 12.21 25.45 ± 12.74 0.11 66.38 ± 68.91 68.32 ± 26.69 0.53

* Indicates statistically significant differenceCNR Contrast-to-noise ratio, FNH focal nodular hyperplasia

B. Homayoon et al.: A retrospective intra-individual crossover comparison

Half-dose (0.05 mmol/kg) gadobenate dimeglumine

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Standard-dose (0.10 mmol/kg) gadodiamide

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Fig. 1. Comparison between half-dose gadobenate dimeg-lumine and standard-dose gadodiamide in a male patient (ageat first scan: 57) with a non-cirrhotic liver demonstrates similar

liver parenchymal and vascular enhancement throughout allphases of contrast enhancement. The MRI examinationswere performed 13 months apart.

Half-dose (0.05 mmol/kg) gadobenate dimeglumine

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Standard-dose (0.10 mmol/kg) gadodiamide

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Fig. 2. Comparison between half-dose gadobenate dimeg-lumine and standard-dose gadodiamide in a female patient(age at first scan: 44) with a single hemangioma in the left he-patic lobe (white arrow) demonstrates similar lesion, liverparenchymal, and vascular enhancement in all phases of

enhancement. The hemangioma in the left hepatic lobe dem-onstrates an early incomplete ring of peripheral puddling andcentripetal progression of contrast enhancement on sub-sequent portal venous and equilibrium phases of enhance-ment. The MRI examinations were performed 19 months apart.

B. Homayoon et al.: A retrospective intra-individual crossover comparison

resonance imaging leads to similar diagnostic resultscompared to standard doses of gadobenate dimeglumineor gadopentetate dimeglumine, for the detection andcharacterization of focal hepatic lesions [15–18]. In aprospective double-blind randomized intra-individualcrossover comparison in 41 patients, Schneider et al.demonstrates that the contrast enhancement and diag-nostic efficacy of half-dose gadobenate dimeglumine areequivalent to standard-dose gadopentetate dimeglumine(0.1 mmol kg-1), for assessment of focal hepatic lesionsduring the dynamic phase of imaging [15]. In a study of47 patients with focal hepatic lesions, with 25 patientsreceiving a standard dose of gadobenate dimeglumineand the remaining 22 receiving a half dose of gadobenatedimeglumine, Nural et al. [18] showed that there were nostatistically significant differences in lesion conspicuityand absolute lesion-liver CNR in patients with benignfocal hepatic lesions or hepatocellular carcinomas whohad received standard-dose and half-dose gadobenatedimeglumine. This study demonstrated a statisticallysignificant difference in CNR of hepatic metastases;however, the comparison was made between groups thathad different primary malignancies. Furthermore, in adouble-blinded clinical trial involving 30 patients, Rungeand Kenney [17] examined the relationship between doseand enhancement characteristics of gadobenate dimeg-

lumine at 0.025, 0.05, 0.10, and 0.20 mmol kg-1 of bodyweight and showed a trend toward improved diagnosticinformation at higher doses, but quantitative differencesbetween scans performed with 0.05 and 0.10 mmol kg-1

of gadobenate dimeglumine did not reach statisticalsignificance in the dynamic phase of imaging.

Our study lends further support to the use of half-dose gadobenate dimeglumine as an alternative to stan-dard doses of extracellular agents, but it is limited in thatit was conducted retrospectively and employs a smallnumber of patients. Those with malignant hepatic lesionswere not included in the study, because they hadundergone interventional or surgical therapy after theinitial MRI examination. We felt that not including thesepatients was necessary as interval therapy may have ledto hemodynamic alterations in the liver, resulting in al-tered signal intensity measurements.

Our study did not include any patients with liver cir-rhosis. Since it is well established that alterations in liverhemodynamics occur with progressive liver fibrosis, fur-ther studies are required to assess contrast enhancementcharacteristics of half-dose gadobenate dimeglumine inpatients with cirrhotic livers [27–29]. Furthermore, ourstudy did not investigate the added benefit of delayedimaging with gadobenate dimeglumine, as this has alreadybeen addressed in prior studies [19].

Half-dose (0.05 mmol/kg) gadobenate dimeglumine

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Standard-dose (0.10 mmol/kg) gadodiamide

20 s post contrast 1 min post contrast 3 min post contrast 5 min post contrast

Fig. 3. Comparison between half-dose gadobenate dimeg-lumine and standard-dose gadodiamide in a female patient(age at first scan: 26) with a single FNH in the left hepatic lobe(white arrow) demonstrates similar lesion, liver parenchymal,

and vascular enhancement. The FNH in the left hepatic lobedemonstrated enhancement in the hepatocyte phase ofimaging on a subsequent study (not shown). The MRIexaminations were performed 19 months apart.

B. Homayoon et al.: A retrospective intra-individual crossover comparison

Conclusion

In conclusion, our study shows that, during the dynamicphase of imaging, a half dose (0.05 mmol kg-1) of gad-obenate dimeglumine results in similar contrastenhancement compared to a standard dose(0.10 mmol kg-1) of gadodiamide, for the assessment ofliver vessels, hemangiomas, and FNHs. Since the utilityof the first generation gadolinium-based contrast agents,such as gadodiamide, is well established in detection andcharacterization of liver lesions and assessment of livervessels [19, 30], by demonstrating similar qualitative andquantitative results between the two contrast agents atthe specified doses, our study suggests that the use ofhalf-dose gadobenate dimeglumine is a reasonablealternative to standard doses of extracellular agents, ifthere is a need to reduce cumulative doses of adminis-tered contrast in patients with renal impairment who areat risk of developing NSF, or if there is a desire to reducethe cost of contrast administration.

The authors hereby certify that this study consists oforiginal unpublished work which is not under consider-ation for publication elsewhere. It complies with all na-tional laws and has received institutional ethics approval.

References

1. Frydrychowicz A, Lubner MG, Brown JJ, et al. (2012) Hepatob-iliary MR imaging with gadolinium-based contrast agents. JMRI35(3):492–511

2. Low RN (2007) Abdominal MRI advances in the detection of livertumours and characterisation. Lancet Oncol 8(6):525–535

3. Lee MS, Lee JY, Kim SH, et al. (2011) Gadoxetic acid disodium-enhanced magnetic resonance imaging for biliary and vascularevaluations in preoperative living liver donors: comparison withgadobenate dimeglumine-enhanced MRI. JMRI 33(1):149–159

4. Filice C, Calliada F, De Masi S, et al. (2011) Italian guidelines fornoninvasive imaging assessment of focal liver lesions: developmentand conclusions. Eur J Gastroenterol Hepatol 23(4):343–353

5. Silva AC, Evans JM, McCullough AE, et al. (2009) MR Imaging ofhypervascular liver masses: a review of current techniques. Radio-Graphics 29:385–403

6. Parente DB, Perez RM, Eiras-Araujo A, et al. (2012) MR imagingof hypervascular lesions in the cirrhotic liver: a diagnostic dilemma.RadioGraphics 32(3):767–787

7. Petersein J, Spinazzi A, Giovagnoni A, et al. (2000) Focal liverlesions: evaluation of the efficacy of gadobenate dimeglumine inMR imaging—a multicenter phase III clinical study. Radiology215:727–736

8. Martin DR, Krishnamoorthy SK, Kalb B, et al. (2010) Decreasedincidence of NSF in patients on dialysis after changing gadoliniumcontrast-enhanced MRI protocols. JMRI 31(2):440–446

9. Perez-Rodriguez J, Lai S, Ehst BD, Fine DM, Bluemke DA (2009)Nephrogenic systemic fibrosis: incidence, associations, and effect ofrisk factor assessment—report of 33 cases. Radiology 250(2):371–377

10. Marckmann P, Skov L, Rossen K, et al. (2007) Case–control studyof gadodiamide related nephrogenic systemic fibrosis. Nephrol DialTransplant 22(11):3174–3178

11. de Campos ROP, Heredia V, Ramalho M, et al. (2011) Quarter-dose (0.025 mmol/kg) gadobenate dimeglumine for abdominalMRI in patients at risk for nephrogenic systemic fibrosis: pre-liminary observations. AJR 196(3):545–552

12. Altun E, Martin DR, Wertman R, et al. (2009) Nephrogenic sys-temic fibrosis: change in incidence following a switch in gadoliniumagents and adoption of a gadolinium policy—report from two USuniversities. Radiology 253(3):689–696

13. Abujudeh HH, Rolls H, Kaewlai R, et al. (2009) Retrospectiveassessment of prevalence of nephrogenic systemic fibrosis (NSF)after implementation of a new guideline for the use of gadobenatedimeglumine as a sole contrast agent for magnetic resonanceexamination in renally impaired patients. JMRI 30:1335–1340

14. Bryant BJ II, Im K, Broome DR (2009) Evaluation of the incidenceof nephrogenic systemic fibrosis in patients with moderate renalinsufficiency administered gadobenate dimeglumine for MRI. ClinRadiol 64(7):706–713

15. Schneider G, Maas R, Kool LS, et al. (2003) Low-dose gadobenatedimeglumine versus standard dose gadopentetate dimeglumine forcontrast-enhanced magnetic resonance imaging of the liver: an in-tra-individual crossover comparison. Invest Radiol 38(2):85–94

16. Kirchin MA, Spinazzi A (2002) Low-dose gadobenate dimeglu-mine-enhanced MRI in the detection and characterization of focalliver lesions. Acad Radiol 9(Suppl 1):121–126

17. Runge VM, Kenney CM (2000) Phase II double-blind, dose-rang-ing clinical evaluation of gadobenate dimeglumine in focal liverlesions: with analysis of liver and kidney signal change on early anddelayed imaging. JMRI 11(6):655–664

18. Nural MS, Gokce E, Danaci M, Bayrak IK, Diren HB (2008) Focalliver lesions: whether a standard dose (0.05 mmol/kg) gadobenatedimeglumine can provide the same diagnostic data as the 0.1 mmol/kg dose. EJR 66:65–74

19. Balci NC, Semelka RC (2005) Contrast agents for MR imaging ofthe liver. Radiol Clin North Am 43(5):887–898

20. Schneider G, Altmeyer K, Kirchin MA, et al. (2007) Evaluation ofa novel time-efficient protocol for gadobenate dimeglumine (Gd-BOPTA)-enhanced liver magnetic resonance imaging. InvestRadiol 42(2):105–115

21. Essig M, Tartaro A, Tartaglione T, et al. (2006) Enhancing lesionsof the brain: intraindividual crossover comparison of contrastenhancement after gadobenate dimeglumine versus establishedgadolinium comparators. Acad Radiol 13(6):744–751

22. Bellin M-F (2006) MR contrast agents, the old and the new. Eur JRadiol 60(3):314–323

23. Maravilla K, Maldjian JA, Schmalfuss I, et al. (2006) Contrastenhancement of central nervous system lesions: multicenter intra-individual crossover comparative study of two MR contrast agents.Radiology 240(2):389–400

24. Essig M (2006) Protocol design for high relaxivity contrast agentsin MR imaging of the CNS. Eur Radiol 16(Suppl 7):M3–M7

25. Wang Y, Alkasab TK, Narin O, et al. (2011) Incidence of neph-rogenic systemic fibrosis after adoption of restrictive gadolinium-based contrast agent guidelines. Radiology 260:105–111

26. Sadowski E, Bennett LK, Chan MR, et al. (2007) Nephrogenicsystemic fibrosis: risk factors and incidence estimation. Radiology243:148–157

27. Lane MJ, Jeffrey RB, Katz DS (2000) Spontaneous intrahepaticvascular shunts. AJR 174(1):125–131

28. Tamada T, Ito K, Higaki A, et al. (2011) Gd-EOB-DTPA-en-hanced MR imaging: evaluation of hepatic enhancement effects innormal and cirrhotic livers. EJR 80(3):e311–e316

29. Faria SC, Ganesan K, Mwangi I, et al. (2009) MR imaging of liverfibrosis: current state of the art. Radiographics 29(6):1615–1635

30. Seale MK, Catalano OA, Saini S, Hahn PF, Sahani DV (2009)Hepatobiliary-specific MR contrast agents: role in imaging the liverand biliary tree. Radiographics 29(6):1725–1748

B. Homayoon et al.: A retrospective intra-individual crossover comparison